| dc.description.abstract | Inter turn short circuit (ITSC) fault is one of the most common faults in permanent magnet
synchronous machines (PMSM). ITSC fault has limited impact on system performance in early
stages, but it can turn into several failures without proper treatment. Therefore, it is important to
characterize ITSC fault of PMSM and develop reliable fault diagnosis algorithms. In the past few
years, many studies proposed different kinds of ITSC fault detection algorithms, such as motor
current signature analysis, parameter identifications, real-time estimators and searching coils.
Recently, the study about ITSC fault mitigation algorithms has drawn more attentions due to its
potential of extending the fault machine lifetime. Instead of immediate replacement, the fault
machine can operate with proper control techniques, which can prevent the deterioration of
existing ITSC fault and reduce the risk of system total failure. This technique is essential to mission
and cost critical systems, which are unable to replace the fault machine immediately. By applying
fault mitigation algorithms, the fault machine can still operate until the next available maintenance
for replacement. However, the precise and reliable mitigation algorithm requires a good
understanding of ITSC fault condition, particularly the number of shorted turns and short circuit
current. The number of shorted turns is the key parameter used in modeling the ITSC fault
machine. The short circuit current is the main reason that causes damage on insulation material. A
reliable mitigation method should be able to limit the short circuit current in a safety range.
However, it is impossible to measure the short circuit current in real application. Therefore, it
requires short circuit current estimation method in order to provide a feedback for evaluating the
performance of ITSC fault mitigation algorithms. In this study, the characterization of ITSC fault
is analyzed in detail at first. Following that, a parameter identification method is provided to
estimate the number of shorted turns at standstill condition. At last, a short circuit current
estimation algorithm is developed based on back electromotive force (back-EMF) estimation. | |
